Contrast extravasation during CT: immediate action, iodine dose and the decision to continue
A 45-year-old woman with type 2 diabetes attends for a contrast-enhanced CT of the abdomen and pelvis to assess a suspected pancreatic mass. Her recent eGFR is 40 mL/min/1.73 m² and she took her usual metformin that morning. The referral notes previous difficult venous access. A 22-gauge cannula is sited in the right antecubital fossa after two attempts. The protocol uses 100 mL of an iodinated contrast medium containing 350 mg iodine per mL, injected at 3 mL/s. At about 30 mL into the injection the patient complains of severe burning at the cannula site. The radiographer stops the injection immediately. There is a firm swelling around the puncture site measuring about 4 cm, the skin is pale, and capillary refill is sluggish, but the radial pulse is normal. The radiologist is in another room reporting. The patient asks whether the test can be finished quickly because she has travelled far.
1. Set out your immediate management of the extravasation, with the reason for each step.
Read the worked answer
Stop and assess. The injection is already stopped. Disconnect the syringe but leave the cannula in place briefly so that any residual contrast can be aspirated back; then remove the cannula. Aspiration reduces the volume left in the tissues.
Limit the injury. Elevate the arm above heart level to encourage drainage and reduce oedema. Apply a cold compress for the first 24 hours, then switch to warm compresses; do not massage the site. Check distal pulses, sensation and capillary refill and mark the edge of the swelling so that progression can be judged objectively.
Escalate. Inform the radiologist and the referring clinician. Most extravasations of this size resolve, but a firm, tensely swollen forearm with sluggish perfusion needs review for compartment syndrome. Provide a written advice sheet and a contact number, and arrange a follow-up call or review at 24 hours.
The patient's travel distance is not a reason to hurry; the priority is to prevent a soft-tissue injury from becoming a compartment syndrome.
2. Calculate how much iodine was extravasated and explain how this informs the decision about completing the study.
Read the worked answer
Calculation. Volume extravasated = about 30 mL. Iodine concentration = 350 mg/mL.
30 mL × 350 mg/mL = 10,500 mg = 10.5 g of iodine.
The planned total was 100 mL × 350 mg/mL = 35 g, so roughly 30% of the intended iodine and 30% of the volume entered the tissues.
Decision. Do not simply reconnect and push the remaining 70 mL through the same fragile 22-gauge cannula. The study is probably non-diagnostic for the pancreas, but the safe course is to discuss with the radiologist whether the images so far answer the question. If they do not, defer the contrast phase for 24–48 hours, use a 20-gauge or larger cannula in the opposite arm, and reduce the flow rate. A 22-gauge cannula in the antecubital fossa is marginal for 3 mL/s; a larger cannula is the main preventive measure for next time.
3. What else must be arranged before she leaves, given her renal function and metformin?
Read the worked answer
Metformin. With an eGFR of 40 mL/min/1.73 m², she is in the group in which metformin is usually withheld at the time of intravascular iodinated contrast and restarted about 48 hours later, after renal function has been rechecked. The extravasation does not change that rule, but the contrast she did receive does trigger it. Contact the referring clinician or diabetes team so the instruction is given by someone who can alter her medication.
Renal risk. An eGFR of 40 is not a contraindication to contrast, but she should be hydrated and her renal function rechecked if there is any concern. Document the baseline eGFR and the volume administered.
Documentation and reporting. Record the cannula size, site, contrast name and concentration, volume extravasated, symptoms, observations, treatment and who was informed. Complete the departmental incident form, inform the radiologist, and ensure the follow-up call is booked. The event should feed into technique review, not just a file.